[llvm] [AMDGPU] Require true low byte extensions in dot4 mul operands (PR #220169)

[via llvm-commits]

github-actions[bot] wrote:

<!--PREMERGE ADVISOR COMMENT: Windows-->
# :window: Windows x64 Test Results

* 141257 tests passed
* 3240 tests skipped
* 1 test failed

## Failed Tests
(click on a test name to see its output)

### LLVM
<details>
<summary>LLVM.CodeGen/AMDGPU/idot4-test.ll</summary>

```
Exit Code: 1

Command Output (stdout):
--
# RUN: at line 2
c:\_work\llvm-project\llvm-project\build\bin\llc.exe -mtriple=amdgpu9.06 < C:\_work\llvm-project\llvm-project\llvm\test\CodeGen\AMDGPU\idot4-test.ll | c:\_work\llvm-project\llvm-project\build\bin\filecheck.exe --check-prefixes=GFX9-DL C:\_work\llvm-project\llvm-project\llvm\test\CodeGen\AMDGPU\idot4-test.ll
# executed command: 'c:\_work\llvm-project\llvm-project\build\bin\llc.exe' -mtriple=amdgpu9.06
# note: command had no output on stdout or stderr
# executed command: 'c:\_work\llvm-project\llvm-project\build\bin\filecheck.exe' --check-prefixes=GFX9-DL 'C:\_work\llvm-project\llvm-project\llvm\test\CodeGen\AMDGPU\idot4-test.ll'
# .---command stderr------------
# | C:\_work\llvm-project\llvm-project\llvm\test\CodeGen\AMDGPU\idot4-test.ll:1245:17: error: GFX9-DL-NEXT: is not on the line after the previous match
# | ; GFX9-DL-NEXT: s_waitcnt lgkmcnt(0)
# |                 ^
# | <stdin>:957:2: note: 'next' match was here
# |  s_waitcnt lgkmcnt(0)
# |  ^
# | <stdin>:955:21: note: previous match ended here
# |  v_mov_b32_e32 v0, 0
# |                     ^
# | <stdin>:956:1: note: non-matching line after previous match is here
# |  v_mov_b32_e32 v3, 0x6050400
# | ^
# | 
# | Input file: <stdin>
# | Check file: C:\_work\llvm-project\llvm-project\llvm\test\CodeGen\AMDGPU\idot4-test.ll
# | 
# | -dump-input=help explains the following input dump.
# | 
# | Input was:
# | <<<<<<
# |              .
# |              .
# |              .
# |            950:  .type dot4_mulhi_bits_not_dot4, at function 
# |            951: dot4_mulhi_bits_not_dot4: ; @dot4_mulhi_bits_not_dot4 
# |            952: ; %bb.0: ; %entry 
# |            953:  s_load_dwordx4 s[0:3], s[4:5], 0x24 
# |            954:  s_load_dwordx2 s[6:7], s[4:5], 0x34 
# |            955:  v_mov_b32_e32 v0, 0 
# | next:1245'0                        {   search range start (exclusive)
# |            956:  v_mov_b32_e32 v3, 0x6050400 
# |            957:  s_waitcnt lgkmcnt(0) 
# | next:1245'1      !~~~~~~~~~~~~~~~~~~~   error: match on wrong line
# |            958:  global_load_ushort v1, v0, s[0:1] 
# |            959:  global_load_ushort v2, v0, s[2:3] 
# |            960:  s_load_dwordx4 s[0:3], s[4:5], 0x3c 
# |            961:  s_waitcnt lgkmcnt(0) 
# |            962:  s_add_i32 s0, s0, s1 
# |              .
# |              .
# |              .
# |           1013:  .set amdgpu.max_num_agpr, 0 
# |           1014:  .set amdgpu.max_num_sgpr, 32 
# |           1015:  .set amdgpu.max_num_named_barrier, 0 
# |           1016:  .section .AMDGPU.csdata,"", at progbits 
# |           1017:  .section ".note.GNU-stack","", at progbits 
# |           1018:  .amd_amdgpu_isa "amdgpu9.06-unknown-unknown-unknown-gfx906" 
# | next:1245'2                                                                  } search range end (exclusive)
# | >>>>>>
# `-----------------------------
# error: command failed with exit status: 1

--

```
</details>

If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label.

https://github.com/llvm/llvm-project/pull/220169